Software-defined service networking: runtime sharing with functional and performance differentiation in multi-tenant cloud applications

Service-oriented architecture offers great flexibility to build, evolve, and deliver composite cloud applications rapidly. With the multi-tenancy principle, the cloud service provider can simultaneously host multiple tenants on the same set of services and their aggregation to improve utilization of services and resources, and to reduce operational cost. However, diverse tenant requirements and dynamic service-oriented environments make supporting multi-tenancy in composite cloud applications challenging. This thesis addresses the problem of how to build and manage composite cloud applications that achieve multi-tenancy in volatile environments, where the capabilities of services and the requirements of the stakeholders are continuously changing.